If you've never seen a data sufficiency question before, the format probably looks strange.
That's a reasonable reaction. Data sufficiency is unique to the GMAT® and related exams like the Executive Assessment. You won't find it on the SAT, the GRE®, or most other standardized tests. So if you're starting GMAT® prep and DS feels confusing, you're in good company.
Here's the thing that trips people up. Data sufficiency questions don't ask you to solve for an answer. They ask whether you have enough information to solve. That's a different skill from regular math problems, and it takes practice to get used to.
The good news is that data sufficiency rewards a consistent process more than almost any other question type on the GMAT®. Once you build that process and make it automatic, DS questions become much more manageable.
This guide covers the format, the two question types, a step-by-step process you can run on every DS question, worked examples, and the strategies that matter most.
Where Data Sufficiency Lives on the GMAT® Focus Edition
On the GMAT® Focus Edition, data sufficiency questions are part of the Data Insights section. They share the section with four integrated reasoning question types: Two-Part Analysis, Graphics Interpretation, Table Analysis, and Multi-Source Reasoning.
The Data Insights section has 20 questions in 45 minutes. You'll see roughly 5 to 8 data sufficiency questions, with the rest being integrated reasoning. The exact mix varies from test to test.
There's a calculator available in Data Insights. You can show it and hide it with a button on the screen.
One change worth noting: about half the DS questions on the Focus Edition are pure logic questions with little or no computation. The process for solving them is the same as classic DS questions, but the math content is lighter.
For a full breakdown of the Data Insights section, see our complete guide to GMAT® Data Insights.
The Data Sufficiency Format
Every data sufficiency question has the same structure.
A question at the top of the screen. Sometimes there's some given information before the question. Sometimes it's just a question by itself.
Two statements below the question. The exam calls them Statement (1) and Statement (2). Think of them as facts. They're always telling you the truth. You can trust them.
Your job is to figure out whether the information in those statements is enough to answer the question definitively.
Not "can I probably figure it out." Not "I think I know the answer." Definitively. One answer. No ambiguity.
The Five Answer Choices
The answer choices are the same on every single data sufficiency question. They never change.
(A) Statement (1) ALONE is sufficient, but Statement (2) alone is not sufficient.
(B) Statement (2) ALONE is sufficient, but Statement (1) alone is not sufficient.
(C) BOTH statements TOGETHER are sufficient, but NEITHER statement ALONE is sufficient.
(D) EACH statement ALONE is sufficient.
(E) Statements (1) and (2) TOGETHER are NOT sufficient.
Memorize these early. You don't want to spend time reading them on test day. Every DS question uses the same five choices, so learning them once pays off on every DS question you see.
What "Sufficient" Means
Sufficient means you can answer the question with one definitive answer.
If the question asks "What is the value of x?" and you can determine that x is always 7, that's sufficient. If x could be 6 in some cases and 7 in other cases, that's not sufficient. You need one answer that works every time.
If the question asks "Is x greater than 5?" and you can determine that the answer is always yes, that's sufficient. If you can determine that the answer is always no, that's also sufficient. A definitive "no" is just as good as a definitive "yes."
But if sometimes the answer is yes and sometimes the answer is no, that's not sufficient. You don't have a definitive answer.
The Two Types of DS Questions
There are two types of data sufficiency questions, and mixing them up is one of the most common DS mistakes we see.
Value Questions
Value questions ask for a specific number. "What is the value of x?" or "How many minutes does it take Andre to complete a task?"
Sufficient data means you get one value. If Andre could take 30 minutes in some cases and 35 minutes in other cases, that's not sufficient. You need to know that it's always one specific number.
Think about value questions like salary questions. If you ask your boss "How much will I make next year?" and they say "Either $50 or $50,000," you'd be confused. You need one number to make a decision. Same thing on value questions. One value is sufficient. Multiple possible values are not.
Yes/No Questions
Yes/no questions ask a question that has a yes or no answer. "Is x equal to 7?" or "Is the perimeter of the rectangle greater than 20?"
Here's the part that can feel tricky. A definitive "yes" is sufficient. A definitive "no" is also sufficient. Either one works. You just need to know for sure.
If the question asks "Is x equal to 7?" and Statement (1) says "x equals 9," that's sufficient. You know for sure the answer is no. It's never 7. That's a definitive answer.
But if Statement (1) says "x is greater than 4," that's not sufficient. x could be 7. x could be 10. Sometimes yes, sometimes no. Not definitive.
Think about yes/no questions like marriage proposals. If you ask someone to marry you and they say yes, that's definitive. You can plan accordingly. If they say no, that's also definitive. You can move on. But if they say "maybe" — sometimes yes, sometimes no — you don't have enough information to make a decision. That's what insufficient data looks like on a yes/no question.
If you find yourself getting confused about which type you're working with, write "VALUE" or "YES/NO" at the top of your scratch pad. It prevents a super common mistake.
The Six-Step Process
Data sufficiency rewards a consistent process more than almost any other question type on the GMAT®. Here's the system we teach, adapted from Season 1, Episode 7 of our podcast series, "Data Sufficiency — What Everyone Should Know."
Step 1: Write What's Given and What's Asked
Physically write it down on your scratch pad. While you're learning the process, do this every time.
If there's a constraint like "x is a positive integer," write it down. If there's no constraint on x, note that too. x could be negative, a fraction, a decimal, zero — anything.
This step sounds obvious. But a lot of us skip it, especially on easier questions. And skipping it's where a lot of DS mistakes come from. The constraints matter. If you forget that x is a positive integer, you might test numbers that aren't allowed. If you forget that x has no constraints, you might assume x is positive when it could be negative.
Writing down what's given and what's asked takes a few seconds. It saves you from missing questions you know how to do.
Step 2: Simplify the Question
Some data sufficiency questions can be simplified with a little algebra. If you can make the question simpler before evaluating the statements, do it.
For example, if the question asks "What is the value of x + y?" and Statement (1) gives you "x + y = 2z and z = 5," you can simplify by substituting: x + y = 2(5) = 10. Now you're looking at a much simpler question.
You won't be able to simplify every question. But when you can, it makes the rest of the problem easier. Build this into your process so you do it automatically.
Step 3: Evaluate Statement (1) Alone
Look at Statement (1) by itself. Don't let Statement (2) cloud your judgment. Pretend it doesn't exist.
Ask: is Statement (1) by itself enough to answer the question?
If Statement (1) is sufficient, you can eliminate B, C, and E. Only A and D are still possible.
If Statement (1) is not sufficient, you can eliminate A and D. Only B, C, and E are still possible.
This elimination step matters. It narrows your choices before you even look at Statement (2), which makes the rest of the question easier.
Step 4: Forget Statement (1). Evaluate Statement (2) Alone
This might be the hardest part of the whole process.
You have to actively forget Statement (1). Scratch it out on your scratch pad if that helps. Then evaluate Statement (2) on its own, in a vacuum.
Why is this so important? Because if you accidentally carry information from Statement (1) into your evaluation of Statement (2), you'll think you have more data than you actually do. That can make an insufficient statement look sufficient. And that's one of the most common ways people get DS questions wrong.
If Statement (2) is sufficient and Statement (1) was also sufficient, pick D.
If Statement (2) is sufficient and Statement (1) was not, pick B.
If Statement (2) is not sufficient, move to Step 5.
Step 5: Combine the Statements (Only If Needed)
You only reach this step if neither statement alone was sufficient. You've already established that Statement (1) by itself isn't enough and Statement (2) by itself isn't enough.
Now you're allowed to put them together.
If combined they're sufficient, pick C.
If combined they're still not enough, pick E.
Don't combine the statements earlier than this. If you combine too early, you create confusion about which statement is doing the work. You might pick C when the answer is really A or B. The wording of C is specific: both statements together are sufficient, but neither alone is sufficient. You can only pick C if you've already ruled out A, B, and D.
Step 6: Pick Your Answer
By this point, the elimination work from Steps 3 through 5 should leave you with one answer. Pick it and move on.
Worked Example 1: A Value Question
Let's walk through a DS question using the six-step process.
What is the value of x?
(1) 2x + 3 = 11
(2) x is a positive integer
Step 1: Write what's given and what's asked.
Asked: the value of x. No constraints in the question stem. x could be anything.
Step 2: Can we simplify the question?
"What is the value of x?" is already as simple as it gets. No simplification needed.
Step 3: Evaluate Statement (1) alone.
2x + 3 = 11. Subtract 3 from both sides: 2x = 8. Divide by 2: x = 4.
One definitive answer. Statement (1) is sufficient. Eliminate B, C, and E.
Step 4: Forget Statement (1). Evaluate Statement (2) alone.
x is a positive integer. So x could be 1, 2, 3, 4, 5, and so on. There's no way to narrow it down to one value.
Statement (2) is not sufficient. Eliminate D.
Step 5: Combine?
Not needed. Statement (1) was already sufficient on its own.
Step 6: Pick the answer.
Statement (1) alone is sufficient. Statement (2) alone is not. That's answer choice (A).
The math here was straightforward. But the point is the process. If you run these six steps on every DS question, you're much less likely to lose track of which statement you're evaluating or what you've already eliminated. On harder questions, that structure is what keeps you from making mistakes.
Worked Example 2: A Yes/No Question
Here's a yes/no question that shows how the process handles a different question type.
Is x greater than 5?
(1) x is greater than 3
(2) x is an even integer greater than 4
Step 1: Write what's given and what's asked.
Asked: is x greater than 5? This is a yes/no question. No constraints on x in the question stem.
Step 2: Can we simplify?
Not really. The question is already simple.
Step 3: Evaluate Statement (1) alone.
x is greater than 3. Could x be greater than 5? Yes — x could be 6, 7, 8. Could x be less than or equal to 5? Also yes — x could be 4, 5, or 3.5.
Sometimes yes, sometimes no. Not definitive. Statement (1) is not sufficient. Eliminate A and D.
Step 4: Forget Statement (1). Evaluate Statement (2) alone.
x is an even integer greater than 4. So x could be 6, 8, 10, 12, and so on.
Is x greater than 5? Every even integer greater than 4 is at least 6. And 6 is greater than 5. So the answer is always yes.
Always yes is a definitive answer. Statement (2) is sufficient. Pick B.
Step 5: Not needed.
Step 6: Answer is (B).
Notice what happened with Statement (1). It was tempting to think "x is greater than 3, so maybe it's greater than 5." But "maybe" isn't good enough on data sufficiency. You need the answer to be the same every time. That's why testing a few numbers matters.
The AD/BCE Elimination Framework
The six-step process gives you a natural elimination system. Here's how it works.
After Step 3 (evaluating Statement (1)):
If Statement (1) is sufficient, the only possible answers are A or D. Think of this as the "AD" branch.
If Statement (1) is not sufficient, the only possible answers are B, C, or E. Think of this as the "BCE" branch.
After Step 4 (evaluating Statement (2)):
In the AD branch: if Statement (2) is also sufficient, pick D. If not, pick A.
In the BCE branch: if Statement (2) is sufficient, pick B. If not, you need to combine (Step 5). If combined they work, pick C. If not, pick E.
This framework keeps you from considering answers that are already ruled out. You can write "AD" or "BCE" at the top of your scratch pad after Step 3 to track where you're at.
Testing Numbers on DS Questions
Testing numbers is one of the most powerful strategies on data sufficiency. You can use it on at least half of DS questions, and for a lot of people, it's the best approach on about a third of DS questions.
The idea is straightforward. Instead of trying to prove sufficiency with algebra, you test specific numbers that fit the statement's constraints. If you find two different numbers that give two different answers to the question, the statement is not sufficient. If every number you test gives the same answer, the statement is probably sufficient.
How to Test Numbers
Pick numbers that fit the constraints in the statement. Then check whether they give the same answer to the question or different answers.
If the statement says "x is a positive integer," test x = 1, x = 2, x = 10. See if the answer to the question changes.
If the statement says "x is an integer" (no "positive" constraint), also test x = 0, x = -1, x = -5. The negative numbers and zero are where most people miss things.
If the statement says "x is a real number," also test fractions and decimals. x = 0.5, x = -2.3, x = 0. These edge cases are where insufficiency hides.
The Key Categories to Test
When you're testing numbers, there are a few categories worth checking:
Positive integers: 1, 2, 10 Zero: 0 Negative integers: -1, -5 Fractions between 0 and 1: 0.5, 0.25 Fractions greater than 1: 1.5, 2.5 Negative fractions: -0.5, -1.5
You won't need all of these on every question. But if the statement allows these types and you're not sure whether the statement is sufficient, testing across these categories will usually reveal it.
When Testing Numbers Proves Insufficiency
If you test two numbers and get two different answers, you've proven the statement is not sufficient. You can stop testing.
For example, if the question is "Is x greater than 5?" and Statement (1) says "x is greater than 3," you test x = 4 (answer: no, 4 is not greater than 5) and x = 6 (answer: yes, 6 is greater than 5). Two different answers. Not sufficient.
When Testing Numbers Suggests Sufficiency
If you test several numbers across different categories and they all give the same answer, that suggests sufficiency. But be careful. Testing numbers can suggest sufficiency without proving it. If you've tested five or six numbers across all the relevant categories and they all agree, the statement is very likely sufficient. But on harder questions, there might be an edge case you didn't think of.
If you can also prove sufficiency with algebra, that's more reliable. Testing numbers is fast and catches most cases. Algebra is slower but catches everything. Use both when you can.
For a deeper dive into the testing-numbers method, see our guide on testing numbers on the GMAT®.
Common DS Traps
Data sufficiency questions are designed to trick you. Here are the most common traps and how to avoid them.
Trap 1: Combining Too Early
This is the C trap. You read both statements, your brain puts them together, and you think "yeah, with both of these I can figure it out." So you pick C.
But maybe one of the statements was sufficient on its own. You didn't check because you combined too fast.
The fix is the six-step process. Evaluate Statement (1) alone first. Then forget it and evaluate Statement (2) alone. Only combine if neither alone was sufficient. If you follow the process, you're much less likely to fall into this trap.
Trap 2: Forgetting Constraints
The question stem says "x is a positive integer." You evaluate the statements and forget that constraint. You test negative numbers. You get the wrong answer.
The fix is Step 1. Write down what's given and what's asked. Every time. If the constraint is on your scratch pad, you're much less likely to forget it.
Trap 3: Assuming No Constraints
The flip side. The question stem says nothing about x. You assume x is positive. You only test positive integers. But x could be negative, zero, or a fraction. The statement might give different answers for those cases.
The fix is the same. Write down what's given. If there are no constraints, note that. "x can be anything" is a constraint in itself. It means you need to test negative numbers, zero, and fractions to be thorough.
Trap 4: A Definitive "No" Is Still Sufficient
On yes/no questions, a lot of people think "no" means insufficient. It doesn't. A definitive "no" is just as sufficient as a definitive "yes."
If the question asks "Is x equal to 5?" and a statement proves x is 12, the answer is no. Always no. That's sufficient. Don't rule out a statement just because it gives you a "no."
Trap 5: "Maybe" Is Not Sufficient
The other side of the same coin. If sometimes the answer is yes and sometimes the answer is no, that's not sufficient. "Maybe" doesn't count.
This trap shows up when a statement narrows things down but doesn't narrow them down enough. x is greater than 3. Is x greater than 5? Maybe. Not sufficient.
Scratch Work Standards
If it's happening in your head on a data sufficiency question, it should be happening on the page.
This goes against what a lot of people will tell you. Many students and even some instructors say you don't need to write everything down on DS. Our experience says otherwise.
The whole point of data sufficiency is to trick you. The questions are designed to make sufficient data look insufficient, or insufficient data look sufficient. Writing out your work helps you track where you're at logically. It frees up brain power for problem-solving instead of holding information in your head.
Your scratch work should be clear enough that someone else could follow your logic. They should be able to look at your scratch pad and tell exactly what your thought process was, step by step, without asking you any questions.
That's a high bar. Most people don't get close to it. But that's the standard worth aiming for.
What to Write on Your Scratch Pad
Here's a template for every DS question:
Write "GIVEN:" and list any constraints from the question stem. If there are none, write "no constraints."
Write "ASKED:" and restate the question. Write "VALUE" or "YES/NO" next to it.
Write "(1)" and evaluate Statement (1). Show your work. Write "SUFF" or "NOT SUFF."
Cross out everything related to Statement (1).
Write "(2)" and evaluate Statement (2). Show your work. Write "SUFF" or "NOT SUFF."
If neither alone was sufficient, write "(1) + (2)" and evaluate them together. Write "SUFF" or "NOT SUFF."
Write your answer choice.
This template takes about 30 seconds once it's a habit. And it prevents the vast majority of DS mistakes.
Timing: Process First, Speed Later
If you're just starting out with data sufficiency, let yourself be slow.
Don't worry about finishing in two minutes. Don't even worry about finishing in five minutes. Focus on building great scratch work habits and stepping through the six-step process on every question.
Speed comes later. Once the process is automatic, you'll start moving faster. But being fast with a bad process isn't much better than being slow with a bad process. Build the habits first.
If you've been studying for a while and you're still struggling with DS, take the time pressure away. Practice untimed. Focus on the process and your scratch work. With enough repetition, you'll build back up to speed with much better habits.
For section-level timing strategy, see our guide on GMAT® timing strategy.
How to Practice DS
Start with official materials. The GMAT® Official Guide has plenty of DS questions. mba.com offers free sample questions. Old GMAT® Classic DS questions are still relevant for Focus Edition prep — the format is the same.
When you review a DS question you missed, ask yourself which step of the process broke down. Did you forget a constraint? Did you combine too early? Did you assume "no" means insufficient? Did you not write anything down?
The pattern of your mistakes tells you what to fix. If you keep forgetting constraints, Step 1 needs more attention. If you keep combining too early, Steps 3 and 4 need more discipline. If you keep misreading yes/no vs. value, the "VALUE" or "YES/NO" label at the top of your scratch pad needs to become automatic.
For a system for tracking and reviewing mistakes, see our guide on the GMAT® error log.
And for when to invest extra time on a DS question versus when to let it go, see our guide on when to let go of a GMAT® question.
FAQ
Is data sufficiency on the GMAT® Focus Edition?
Yes. Data sufficiency is part of the Data Insights section on the GMAT® Focus Edition. You'll see roughly 5 to 8 DS questions out of 20 Data Insights questions. The format is the same as it was on the classic GMAT®. About half the DS questions on the Focus Edition are pure logic with little or no computation.
Is data sufficiency part of the Quant section on the GMAT® Focus Edition?
No. On the Focus Edition, data sufficiency moved from the Quantitative section to the Data Insights section. The Quantitative section is now Problem Solving only. Data Insights combines DS with four integrated reasoning question types.
Are the DS answer choices the same on every question?
Yes. The five answer choices (A through E) are identical on every data sufficiency question. Memorize them so you don't spend time reading them on test day.
What is the difference between a value question and a yes/no question on DS?
Value questions ask for a specific number. "What is the value of x?" Sufficient data means one definitive value. Yes/no questions ask a question with a yes or no answer. "Is x greater than 5?" A definitive "yes" is sufficient. A definitive "no" is also sufficient. "Sometimes yes, sometimes no" is not sufficient.
Can a "no" answer be sufficient on data sufficiency?
Yes. On a yes/no question, a definitive "no" is just as sufficient as a definitive "yes." If the question asks "Is x equal to 5?" and a statement proves x is always 9, the answer is always no. That's sufficient. The key is definitiveness, not which answer you get.
Should you always test numbers on data sufficiency?
Testing numbers is a powerful strategy on many DS questions, but it isn't the only approach. For about half of DS questions, testing numbers can help you determine sufficiency. For the other half, algebraic reasoning may be more efficient. Build both skills. Use testing numbers when the algebra isn't obvious, and use algebra when you can prove sufficiency directly.
How long should you spend on a data sufficiency question?
Most DS questions take about 2 minutes once you're comfortable with the process. But when you're learning, let yourself take 3 to 5 minutes per question. Build the process first. Speed follows once the process becomes automatic.
Should you combine the statements as soon as possible?
No. Only combine the statements after you've evaluated each one alone and determined that neither is sufficient by itself. Combining too early is one of the most common DS mistakes. It can make you pick C when the answer is really A, B, or D.
Do you need to write everything down on DS?
We recommend writing out your work on every DS question. Your scratch work should be clear enough that someone else could follow your logic without asking you any questions. This takes practice, but it prevents the majority of DS mistakes — which usually come from losing track of which statement you're evaluating or forgetting a constraint.
Are old GMAT® Classic data sufficiency materials still useful?
Yes. The data sufficiency format hasn't changed on the Focus Edition. All DS questions from GMAT® Classic prep materials are still relevant. The only difference is that DS now lives in the Data Insights section instead of the Quant section, and about half the Focus Edition DS questions are pure logic with less computation. But the process for solving them is the same.
Want to learn even more?
We covered this entire process in depth on the podcast. Season 1, Episode 7 of our podcast series, "Data Sufficiency — What Everyone Should Know," walks through the format, the two question types, the six-step process, testing numbers, and scratch work habits in detail.
You can listen to the full episode on Spotify, Apple Podcasts, or on our podcast page.
If you're building your study plan, here are a few related guides:
- GMAT® Data Insights: A Complete Guide
- How to Study for the GMAT®: A Complete Guide
- Testing Numbers on the GMAT®: A Simple System
- GMAT® Timing Strategy: How to Pace Each Section
- GMAT® Error Log: How to Track and Review Mistakes
- GMAT® Practice Test Strategy: How Many to Take and When
- How to Break Through a GMAT® Score Plateau
- When to Let Go of a GMAT® Question
- GMAT® Number Properties: A Complete Guide
- Best GMAT® Study Materials and Resources for 2026